CN111223099B - Contraband detection method, computing equipment and storage medium - Google Patents
Contraband detection method, computing equipment and storage medium Download PDFInfo
- Publication number
- CN111223099B CN111223099B CN202010302804.2A CN202010302804A CN111223099B CN 111223099 B CN111223099 B CN 111223099B CN 202010302804 A CN202010302804 A CN 202010302804A CN 111223099 B CN111223099 B CN 111223099B
- Authority
- CN
- China
- Prior art keywords
- detection
- contraband
- iou
- box
- score
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Active
Links
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING OR CALCULATING; COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/0002—Inspection of images, e.g. flaw detection
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01V—GEOPHYSICS; GRAVITATIONAL MEASUREMENTS; DETECTING MASSES OR OBJECTS; TAGS
- G01V5/00—Prospecting or detecting by the use of ionising radiation, e.g. of natural or induced radioactivity
- G01V5/20—Detecting prohibited goods, e.g. weapons, explosives, hazardous substances, contraband or smuggled objects
- G01V5/22—Active interrogation, i.e. by irradiating objects or goods using external radiation sources, e.g. using gamma rays or cosmic rays
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01V—GEOPHYSICS; GRAVITATIONAL MEASUREMENTS; DETECTING MASSES OR OBJECTS; TAGS
- G01V8/00—Prospecting or detecting by optical means
- G01V8/10—Detecting, e.g. by using light barriers
-
- G—PHYSICS
- G06—COMPUTING OR CALCULATING; COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T2207/00—Indexing scheme for image analysis or image enhancement
- G06T2207/20—Special algorithmic details
- G06T2207/20081—Training; Learning
-
- G—PHYSICS
- G06—COMPUTING OR CALCULATING; COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T2207/00—Indexing scheme for image analysis or image enhancement
- G06T2207/20—Special algorithmic details
- G06T2207/20084—Artificial neural networks [ANN]
-
- G—PHYSICS
- G06—COMPUTING OR CALCULATING; COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T2207/00—Indexing scheme for image analysis or image enhancement
- G06T2207/30—Subject of image; Context of image processing
- G06T2207/30204—Marker
Landscapes
- Physics & Mathematics (AREA)
- Engineering & Computer Science (AREA)
- General Physics & Mathematics (AREA)
- Life Sciences & Earth Sciences (AREA)
- General Life Sciences & Earth Sciences (AREA)
- Geophysics (AREA)
- Quality & Reliability (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Theoretical Computer Science (AREA)
- High Energy & Nuclear Physics (AREA)
- Image Analysis (AREA)
Abstract
The invention relates to a contraband detection method, a computing device and a storage medium, wherein the detection method determines a detection result by judging whether a detection frame with the IOU (input output) of the detection frame with the highest prediction score is larger than a filtering threshold value exists in the detection results of the same prediction category. And when the IOU is larger than the background threshold, the detection result is the integrated result of the detection frame with the highest prediction score and the detection frame corresponding to the largest IOU in the rest detection frames, and when the IOU is not larger than the filtering threshold, the detection result of the detection frame with the highest prediction score is reserved. The integration method is simple and effective, does not need to set different judgment parameters aiming at different detection models, has universality in each model, is flexible to apply, consumes less computing resources and has high model operation speed. The defects of low contraband detection rate, high false detection rate and the like in a complex scene in the prior art are effectively overcome, the robustness of the detection model is enhanced, and the contraband detection performance of the model in the complex scene is improved.
Description
Technical Field
The invention belongs to the technical field of security inspection, and particularly relates to a contraband detection method, a computing device and a storage medium.
Background
The detection of contraband articles in X-ray pictures is a key problem in the field of public security, and the main purpose of the detection is to accurately detect whether the X-ray pictures contain the contraband articles such as knives and guns in the X-ray pictures after pseudo-color processing, and to position the contraband articles under the condition of complex shielding interference, thereby providing clues for manual unpacking inspection. At present, the contraband detection in the X-ray picture is mainly achieved through manual inspection, related workers need to monitor pictures all the time and give out manual judgment results in a short time, time and labor are wasted, the requirements for high speed and large quantity in express logistics security inspection cannot be met completely, and if the contraband is omitted through manual inspection, serious consequences can be caused. Therefore, the automated contraband detection technology has received much attention from people due to its advantages of low cost, high efficiency, and high stability.
Object Detection (Object Detection) is one of basic tasks in the field of computer vision, in recent years, with the rise of deep learning, more and more Object Detection algorithms are completed by adopting a deep learning method, a large number of algorithms and technologies based on deep learning are provided, and the Detection precision of the contraband Detection task is continuously refreshed on a plurality of public data sets. Nevertheless, for a data set with a complex scene or video data close to a living scene, due to the influence of factors such as illumination change, complex background, and difference in viewing angle, most algorithms cannot achieve a satisfactory detection effect in such a scene, and have a certain distance from large-scale commercial application, so the research on the current security detection technology of contraband still has a challenge.
Disclosure of Invention
The present invention is directed to overcome the disadvantages of the prior art, and provides a contraband detection method, a computing device and a storage medium. The method effectively overcomes the defects of low contraband detection rate, high false detection rate and the like in a complex scene in the prior art, enhances the robustness of the detection model, and improves the contraband detection performance of the model in the complex scene.
To achieve the above object, according to a first aspect of the present invention, a contraband detection method is provided, which includes the following steps:
s1, respectively detecting the security inspection image to be detected by using N preset different contraband detection models to obtain N detection results; the detection result comprises detection frames of contraband on the image and a prediction category and a prediction score corresponding to each detection frame; wherein N is an integer not less than 2;
s2, screening the detection results in the same prediction category, if IOU is existed, retaining the detection frame with the highest prediction score and the detection frame corresponding to the largest IOU in the rest detection frames, and retaining the integration result after integrating the detection frames; if the IOU does not exist, retaining the detection result of the detection box with the highest prediction score; the IOU is the coincidence degree of the detection frame with the highest prediction score and the residual detection frames in each prediction category for the set filtering threshold value;
the integration method comprises the following steps: let the detection Box with the highest prediction score be Boxx，BoxxThe coordinate is [ X1 ]x, Y1x, X2x,Y2x]Predicted score of Cx(ii) a The detection Box corresponding to the largest IOU in the rest detection boxes is Boxy，BoxyHas the coordinate of [ X1 ]y,Y1y, X2y, Y2y]Predicted score of Cy(ii) a Integrated detection frame coordinatesComprises the following steps:
the score after integration is C:
；
s3 outputs the detection result.
According to a further aspect of the present invention, there is provided a computing device comprising a memory, a processor and computer instructions stored on the memory and executable on the processor, the processor implementing the steps of the detection method as described above when executing the instructions.
According to yet another aspect of the present invention, a computer readable storage medium is provided, which stores computer instructions that, when executed by a processor, implement the steps of the detection method as described above.
The invention has the beneficial technical effects that: because the detection accuracy of a single model is very limited, the invention uses a multi-model integration method to learn the contribution of each model to the final result and improve the sensitivity of the model, and is different from the conventional multi-model integration thought. And when the IOU is larger than the filtering threshold value, the detection result is the integrated result of the detection frame with the highest prediction score and the detection frame corresponding to the largest IOU in the rest detection frames, and when the IOU is not larger than the filtering threshold value, the detection result of the detection frame with the highest prediction score is reserved. The integration method is simple and effective, does not need to set different judgment parameters aiming at different detection models, has universality in each model, is flexible to apply, consumes less computing resources and has high model operation speed. The defects of low contraband detection rate, high false detection rate and the like in a complex scene in the prior art are effectively overcome, the robustness of the detection model is enhanced, and the contraband detection performance of the model in the complex scene is improved. Further, the integrated detection method of the present invention also enables us to compare different model architectures to continuously improve our products.
Drawings
FIG. 1 is a block diagram of a computing device provided by an embodiment of the present specification;
fig. 2 is a flowchart of a method for detecting contraband in a complex scenario according to an embodiment of the present disclosure;
fig. 3 is a flowchart of a method for detecting contraband according to an embodiment of the present disclosure.
Detailed Description
In the following description, numerous specific details are set forth in order to provide a thorough understanding of the present application. This application is capable of implementation in many different ways than those herein set forth and of similar import by those skilled in the art without departing from the spirit of this application and is therefore not limited to the specific implementations disclosed below.
The terminology used in the description of the one or more embodiments is for the purpose of describing the particular embodiments only and is not intended to be limiting of the description of the one or more embodiments. As used in one or more embodiments of the present specification and the appended claims, the singular forms "a," "an," and "the" are intended to include the plural forms as well, unless the context clearly indicates otherwise. It should also be understood that the term "and/or" as used in one or more embodiments of the present specification refers to and encompasses any and all possible combinations of one or more of the associated listed items.
It will be understood that, although the terms first, second, etc. may be used herein in one or more embodiments to describe various information, these information should not be limited by these terms. These terms are only used to distinguish one type of information from another. For example, a first can also be referred to as a second and, similarly, a second can also be referred to as a first without departing from the scope of one or more embodiments of the present description. The word "if" as used herein may be interpreted as "at … …" or "when … …" or "in response to a determination", depending on the context.
First, the noun terms to which one or more embodiments of the present invention relate are explained.
Contraband: articles that are not legally required to be manufactured, purchased, used, held, stored, transported in and out of the mouth, such as weapons, ammunition, explosive articles (e.g., explosives, detonators, fuse cords, etc.), and the like.
And (4) security inspection images: the security inspection equipment or security inspection machine related to the invention is not limited to the X-ray security inspection equipment or the specific type or model of the security inspection equipment by using the image acquired by the security inspection equipment, and the security inspection equipment and/or security inspection machine which can be realized by scanning are both the protection scope of the invention, such as terahertz imaging equipment and the like.
In the present application, a contraband detection method, a computing device and a storage medium are provided, which are described in detail one by one in the following embodiments.
FIG. 1 shows a block diagram of a computing device 100, according to an embodiment of the present description. The components of the computing device 100 include, but are not limited to, memory 110 and processor 120. The processor 120 is coupled to the memory 110 via a bus 130 and a database 150 is used to store data.
The computing device 100 also includes AN access device 140, the access device 140 enabling the computing device 100 to communicate via one or more networks 160, examples of which include a Public Switched Telephone Network (PSTN), a local area network (L AN), a Wide Area Network (WAN), a Personal Area Network (PAN), or a combination of communication networks such as the Internet the access device 140 may include one or more of any type of network interface, e.g., a Network Interface Card (NIC), wired or wireless, such as AN IEEE802.11 Wireless local area network (W L AN) wireless interface, a Global microwave Internet Access (Wi-MAX) interface, AN Ethernet interface, a Universal Serial Bus (USB) interface, a cellular network interface, a Bluetooth interface, a Near Field Communication (NFC) interface, and so forth.
In one embodiment of the present description, the above-described components of computing device 100 and other components not shown in FIG. 1 may also be connected to each other, such as by a bus. It should be understood that the block diagram of the computing device architecture shown in FIG. 1 is for purposes of example only and is not limiting as to the scope of the description. Those skilled in the art may add or replace other components as desired.
Computing device 100 may be any type of stationary or mobile computing device, including a mobile computer or mobile computing device (e.g., tablet, personal digital assistant, laptop, notebook, netbook, etc.), a mobile phone (e.g., smartphone), a wearable computing device (e.g., smartwatch, smartglasses, etc.), or other type of mobile device, or a stationary computing device such as a desktop computer or PC. Computing device 100 may also be a mobile or stationary server.
Wherein the processor 120 may perform the steps of the method shown in fig. 2. Fig. 2 is a flowchart illustrating a method for detecting contraband in a complex scenario according to an embodiment of the present disclosure, which includes the following steps.
S1: and acquiring security inspection images of the articles and preprocessing the images to form a data set.
The above items include, but are not limited to, packages, luggage, bags, daily items, and contraband, and are specifically determined according to the needs of the actual application scenario, and the scenario may be, for example, an airport, a train station, a bus station, a government building, a embassy, a conference center, a convention and exhibition center, a hotel, a market, a major event, a post office, a school, a logistics industry, industrial inspection, and an express transit station. The main article that different scenes correspond is different, for example this kind of transportation scene of airport railway station uses luggage, package, bag as main safety inspection article more, and express delivery parcel, express delivery case, container are main safety inspection article in logistics industry, express delivery transfer. Preferably, the more security images the better. The data are preprocessed in a mode of preprocessing a format which can be read by a neural network, normalizing, denoising, background difference, artifact removing and data enhancement, and the training model has better robustness through means of geometric transformation operation and/or pixel transformation operation and the like.
The image is normalized by a predetermined size, for example 500 × 500 in this embodiment.
Denoising the image by using a Gaussian smoothing algorithm, wherein the value of each point of the image after the Gaussian smoothing is obtained by weighting and averaging the value of each point and other pixel values in the field; the specific operation is to scan each pixel in the image by using a template, and replace the value of the central pixel point of the template by the weighted average gray value of the pixels in the field determined by the template.
After Gaussian smoothing, fine noise on the image is removed, and although edge information in the image is weakened to a certain extent, edges are still reserved relative to noise.
The background difference algorithm extracts the gray value median of the whole image (500 × 500) as the gray value of the background, and then calculates the difference absolute value between the gray value of each pixel point in the image and the background: i issub＝|Ifg-bg |, where bg is the median of the whole image, IfgFor each pixel gray value in the image, it is known that the foreign object points have larger difference points than the difference between the background point and the background gray value, so the absolute value I of the difference value issubRegarding the probability that a pixel belongs to a foreign object point, the larger the value, the more likely the corresponding pixel is to be a foreign object point.
S2: and respectively detecting the images in the data set in the S1 by utilizing preset N different contraband detection models, detecting the contraband in the images, and marking a detection frame on the contraband, wherein N is an integer more than or equal to 2.
Specifically, the contraband detection model is a convolutional neural network model, and is preferably selected from two-stage detection models such as Cascade-RCNN, fast-RCNN, grid RCNN, efficientDet, L ibra RCNN and the like.
As an example of one embodiment, as shown in fig. 1, the method is introduced by taking a case where the number N of the preset contraband detection models is 5 as an example, the preset contraband detection models are contraband detection models N1, N2, N3, N4, and N5, which are different from each other and are obtained by pre-training.
Preferably, the preset different contraband detection models have the same training process, and the method comprises the following steps:
s21: collecting security inspection images containing contraband, acquiring an image set and corresponding target labels, and constructing a training data set, wherein the method specifically comprises the following steps:
i. collecting security images containing contraband (such as handguns, lighters, etc.) as a sample set;
ii, manually marking the sample set by using a marking tool, wherein the manually marked content comprises the position and the category of the contraband;
performing data enhancement processing on the sample set, wherein the data enhancement processing method comprises the steps of performing any one or combination of any several of rotation, translation, turnover, contrast change and Gaussian noise on the security inspection image, performing synchronous data enhancement operation on a target label in the security inspection image while performing data enhancement processing on the security inspection image, and increasing the sample data volume after data enhancement by the enhancement of the target label to jointly form a training data set;
s22: and respectively training the contraband detection models N1-N5 by using the training data sets in the S21, so that each model obtains an optimal result, and the preset 5 different contraband detection models are obtained.
S3, the same image is detected by the preset N different contraband detection models in S2, and the obtained detection frame is divided into different sets according to the prediction categories.
Specifically, the same image is detected by different contraband detection models to obtain N detection results, wherein the detection results include detection frames for contraband, and prediction categories and prediction scores corresponding to the detection frames. For example, the detection result of the pistol is set G1 for the prediction category, G2 for the lighter for the prediction category, and G3 for the battery for the prediction category.
Preferably, the descending sequence list is obtained by descending order in each set according to the prediction scores of the detection frames.
S4 sets the filtering threshold value to 0 < 1, and calculates the IOU of the detection box with the highest predictive score and the IOU of the remaining detection boxes in the descending list for each set.
If the IOU is higher, retaining the detection frame with the highest prediction score and the detection frame corresponding to the maximum IOU in the rest detection frames, and integrating the detection frames; the integration method comprises the following steps: the detection Box with the highest prediction score is marked as BoxxGet BoxxCoordinate of (2) [ X1x,Y1x,X2x,Y2x]And a prediction score Cx(ii) a The detection Box corresponding to the maximum IOU in the rest detection boxes is marked as Boxy，BoxyHas the coordinate of [ X1 ]y,Y1y,X2y,Y2y]Predicted score of Cy(ii) a The coordinates of the integrated detection frame are as follows:
the score after integration is C:
；
if IOU does not exist, retaining the detection Box Box with the highest prediction scorexAnd its coordinates [ X1 ]x,Y1x,X2x,Y2x]And its predicted score Cx。
The degree of coincidence of the two detection frames, i.e. the ratio of the intersection and the union of the two detection frames, is evaluated by the IOU value, and is used for screening and filtering the detection results, and the value is an empirical value, and is usually set to 0.5.
S5 iterates S3, S4 for a number of times until all resulting images in S2 are screened and integrated.
S6 outputs the detection result.
Because the detection accuracy of a single model is very limited, the embodiment of the invention uses a multi-model integration method to learn the contribution of each model to the final result and improve the sensitivity of the model, and is different from the conventional thought of multi-model integration. And when the IOU is larger than the filtering threshold value, the detection result is the integrated result of the detection frame with the highest prediction score and the detection frame corresponding to the largest IOU in the rest detection frames, and when the IOU is not larger than the filtering threshold value, the detection result of the detection frame with the highest prediction score is reserved. Compared with the prior art, for example, patent inventions with application numbers of 201910926982.X and the like, on the premise of not reducing the detection effect, different judgment parameters do not need to be set for different detection models, and the method has the advantages of universality, flexible application, less consumption of computing resources and high model operation speed in each model. The defects of low contraband detection rate, high false detection rate and the like in a complex scene in the prior art are effectively overcome, the robustness of the detection model is enhanced, and the contraband detection performance of the model in the complex scene is improved. Further, the integrated detection method of the present invention also enables us to compare model architectures of different combinations to continuously improve our products.
Fig. 3 illustrates a method for detecting contraband according to another embodiment of the present disclosure, which is used for security image detection without considering a detection scenario of contraband, for example, in an express transit yard of a certain district level, express packages are different in size and various in variety, and the size and the variety of the detected image of an express package cannot find a rule every day; or the detection scenario is not counted. The method comprises the following steps:
s1, respectively detecting the security inspection image to be detected by using N preset different contraband detection models to obtain N detection results; the detection result comprises detection frames of contraband on the image and a prediction category and a prediction score corresponding to each detection frame; wherein N is an integer not less than 2;
s2, screening the detection results in the same prediction category, if IOU is existed, retaining the detection frame with the highest prediction score and the detection frame corresponding to the largest IOU in the rest detection frames, and retaining the integration result after integrating the detection frames; if the IOU does not exist, retaining the detection result of the detection box with the highest prediction score; wherein, the set filtering threshold is an empirical value, which is usually set to 0.5, and the IOU is the coincidence degree of the detection frame with the highest prediction score in each prediction category and the remaining detection frames; the integration method comprises the following steps: let the detection Box with the highest prediction score be Boxx，BoxxThe coordinate is [ X1 ]x,Y1x,X2x,Y2x]Predicted score of Cx(ii) a The detection Box corresponding to the largest IOU in the rest detection boxes is Boxy，BoxyHas the coordinate of [ X1 ]y,Y1y,X2y,Y2y]Predicted score of Cy(ii) a The coordinates of the integrated detection frame are as follows:
the score after integration is C:
；
s3 outputs the detection result.
An embodiment of the present application further provides a computer readable storage medium, which stores computer instructions, and when the instructions are executed by a processor, the instructions implement the steps of the detection method as described above.
The above is an illustrative scheme of a computer-readable storage medium of the present embodiment. It should be noted that the technical solution of the storage medium belongs to the same concept as the technical solution of the detection method, and details that are not described in detail in the technical solution of the storage medium can be referred to the description of the technical solution of the detection method.
The foregoing description has been directed to specific embodiments of this disclosure. Other embodiments are within the scope of the following claims. In some cases, the actions or steps recited in the claims may be performed in a different order than in the embodiments and still achieve desirable results. In addition, the processes depicted in the accompanying figures do not necessarily require the particular order shown, or sequential order, to achieve desirable results. In some embodiments, multitasking and parallel processing may also be possible or may be advantageous.
The computer instructions comprise computer program code which may be in the form of source code, object code, an executable file or some intermediate form, or the like. The computer-readable medium may include: any entity or system capable of carrying said computer program code, recording medium, usb disk, removable hard disk, magnetic disk, optical disk, computer Memory, Read-Only Memory (ROM), Random Access Memory (RAM), electrical carrier wave signals, telecommunications signals, software distribution medium, etc. It should be noted that the computer readable medium may contain content that is subject to appropriate increase or decrease as required by legislation and patent practice in jurisdictions, for example, in some jurisdictions, computer readable media does not include electrical carrier signals and telecommunications signals as is required by legislation and patent practice.
It should be noted that, for the sake of simplicity, the above-mentioned method embodiments are described as a series of acts or combinations, but those skilled in the art should understand that the present application is not limited by the described order of acts, as some steps may be performed in other orders or simultaneously according to the present application. Further, those skilled in the art should also appreciate that the embodiments described in the specification are preferred embodiments and that the acts and modules referred to are not necessarily required in this application.
In the above embodiments, the descriptions of the respective embodiments have respective emphasis, and for parts that are not described in detail in a certain embodiment, reference may be made to related descriptions of other embodiments.
The preferred embodiments of the present application disclosed above are intended only to aid in the explanation of the application. Alternative embodiments are not exhaustive and do not limit the invention to the precise embodiments described. Obviously, many modifications and variations are possible in light of the above teaching. The embodiments were chosen and described in order to best explain the principles of the application and the practical application, to thereby enable others skilled in the art to best understand and utilize the application. The application is limited only by the claims and their full scope and equivalents.
Claims (9)
1. A method for contraband detection, comprising the steps of:
s1, respectively detecting the security inspection image to be detected by using N preset different contraband detection models to obtain N detection results; the detection result comprises detection frames of contraband on the image and a prediction category and a prediction score corresponding to each detection frame; wherein N is an integer not less than 2;
s2, screening the detection results in the same prediction category, if IOU is existed, retaining the detection frame with the highest prediction score and the detection frame corresponding to the largest IOU in the rest detection frames, and retaining the integration result after integrating the detection frames; if the IOU does not exist, retaining the detection result of the detection box with the highest prediction score; the IOU is the coincidence degree of the detection frame with the highest prediction score and the residual detection frames in each prediction category for the set filtering threshold value;
the integration method comprises the following steps: let the detection Box with the highest prediction score be Boxx，BoxxThe coordinate is [ X1 ]x,Y1x,X2x,Y2x]Predicted score of Cx(ii) a The detection Box corresponding to the largest IOU in the rest detection boxes is Boxy，BoxyHas the coordinate of [ X1 ]y,Y1y,X2y,Y2y]Predicted score of Cy(ii) a The coordinates of the integrated detection frame are as follows:
the score after integration is C:
；
s3 outputs the detection result.
2. A contraband detection method under a complex scene is characterized by comprising the following steps:
s1: acquiring security inspection images of articles and preprocessing the images to form a data set;
s2: respectively detecting the images in the data set in S1 by using preset N different contraband detection models, detecting contraband in the images, and marking a detection frame on the contraband, wherein N is an integer more than or equal to 2;
s3, dividing a detection frame obtained after the same image is detected by the preset N different contraband detection models in S2 into different sets according to prediction categories;
specifically, the same image is detected by different contraband detection models to obtain N detection results, wherein the detection results comprise detection frames for contraband and prediction categories and prediction scores corresponding to the detection frames;
s4, setting a filtering threshold value to be 0 < 1, and respectively calculating the IOU of the detection frame with the highest prediction score and the IOU of the residual detection frames in each set, wherein the IOU is the coincidence degree of the detection frame with the highest prediction score and the residual detection frames in each prediction category;
if the IOU is higher, retaining the detection frame with the highest prediction score and the detection frame corresponding to the maximum IOU in the rest detection frames, and integrating the detection frames; the integration method comprises the following steps: the detection Box with the highest prediction score is marked as BoxxGet BoxxCoordinate of (2) [ X1x,Y1x,X2x,Y2x]And a prediction score Cx(ii) a The detection Box corresponding to the maximum IOU in the rest detection boxes is marked as Boxy，BoxyHas the coordinate of [ X1 ]y,Y1y,X2y,Y2y]Predicted score of Cy(ii) a The coordinates of the integrated detection frame are as follows:
the score after integration is C:
；
if IOU does not exist, retaining the detection Box Box with the highest prediction scorexAnd its coordinates [ X1 ]x,Y1x,X2x,Y2x]And its predicted score Cx；
S5, iterating S3 and S4 for multiple times until all the obtained images in S2 are screened and integrated;
s6 outputs the detection result.
3. The method for detecting contraband in a complex scene according to claim 2, wherein the preset different contraband detection models have the same training process, and the method comprises the following steps:
s21: collecting security inspection images containing contraband, acquiring an image set and corresponding target labels, and constructing a training data set, wherein the method specifically comprises the following steps:
i. collecting a security inspection image containing contraband as a sample set;
ii, manually marking the sample set by using a marking tool, wherein the manually marked content comprises the position and the category of the contraband;
performing data enhancement processing on the sample set, wherein the data enhancement processing method comprises the steps of performing any one or combination of any several of rotation, translation, turnover, contrast change and Gaussian noise on the security inspection image, performing data enhancement processing on the security inspection image, and simultaneously performing synchronous data enhancement operation on a target label in the security inspection image, so that the sample data size after data enhancement is increased to jointly form a training data set;
s22: and training the contraband detection models by using the training data set in the S21 respectively to obtain the optimal result of each model, so as to obtain the preset N different contraband detection models.
4. The method according to claim 2, wherein the detection of contraband in a complex scene is performed in each set according to the prediction scores of the detection frames in a descending order to obtain a corresponding descending sequence table.
5. The method for detecting contraband in a complex scene according to claim 2, wherein the method is an empirical value.
6. The method for detecting contraband in a complex scene according to claim 5, wherein the value is 0.5.
7. The method for detecting contraband in a complex scene according to claim 2, wherein the degree of coincidence is a ratio of an intersection and a union.
8. A computing device comprising a memory, a processor, and computer instructions stored on the memory and executable on the processor, wherein the processor implements the steps of the contraband detection method of any of claims 1-7 when executing the instructions.
9. A computer-readable storage medium storing computer instructions which, when executed by a processor, perform the steps of the contraband detection method according to any of claims 1 to 7.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202010302804.2A CN111223099B (en) | 2020-04-17 | 2020-04-17 | Contraband detection method, computing equipment and storage medium |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202010302804.2A CN111223099B (en) | 2020-04-17 | 2020-04-17 | Contraband detection method, computing equipment and storage medium |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN111223099A CN111223099A (en) | 2020-06-02 |
| CN111223099B true CN111223099B (en) | 2020-07-28 |
Family
ID=70808389
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202010302804.2A Active CN111223099B (en) | 2020-04-17 | 2020-04-17 | Contraband detection method, computing equipment and storage medium |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN111223099B (en) |
Families Citing this family (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN111563494B (en) * | 2020-07-16 | 2020-10-27 | 平安国际智慧城市科技股份有限公司 | Behavior identification method and device based on target detection and computer equipment |
| WO2022016355A1 (en) * | 2020-07-21 | 2022-01-27 | 华为技术有限公司 | Method and apparatus for determining labeling information |
| CN112257752A (en) * | 2020-09-23 | 2021-01-22 | 上海东普信息科技有限公司 | Express item identification and classification method, device, equipment and storage medium |
| CN113920285B (en) * | 2021-09-03 | 2025-05-23 | 浙江大华技术股份有限公司 | Object detection method, device, electronic device and storage medium |
| CN114693968A (en) * | 2022-03-23 | 2022-07-01 | 成都智元汇信息技术股份有限公司 | Verification method and system based on intelligent image recognition box performance |
Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110428007A (en) * | 2019-08-01 | 2019-11-08 | 科大讯飞(苏州)科技有限公司 | X-ray image object detection method, device and equipment |
| US10665018B2 (en) * | 2014-04-18 | 2020-05-26 | Magic Leap, Inc. | Reducing stresses in the passable world model in augmented or virtual reality systems |
-
2020
- 2020-04-17 CN CN202010302804.2A patent/CN111223099B/en active Active
Patent Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US10665018B2 (en) * | 2014-04-18 | 2020-05-26 | Magic Leap, Inc. | Reducing stresses in the passable world model in augmented or virtual reality systems |
| CN110428007A (en) * | 2019-08-01 | 2019-11-08 | 科大讯飞(苏州)科技有限公司 | X-ray image object detection method, device and equipment |
Non-Patent Citations (2)
| Title |
|---|
| 改进的非极大值抑制算法的目标检测;赵文清等;《中国图象图形学报》;20181130;第23卷(第11期);全文 * |
| 联合特征相似性度量和交并比的检测框优选研究;孙世强等;《电脑知识与技术》;20191031;第15卷(第29期);全文 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN111223099A (en) | 2020-06-02 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN111223099B (en) | Contraband detection method, computing equipment and storage medium | |
| CN110956225B (en) | Contraband detection method and system, computing device and storage medium | |
| CN110765990A (en) | Intelligent article detection method and system, computing device and storage medium | |
| US10943145B2 (en) | Image processing methods and apparatus, and electronic devices | |
| EP3985620B1 (en) | Fine-grained classification of retail products | |
| Zhang et al. | Toward new retail: A benchmark dataset for smart unmanned vending machines | |
| CN108427927B (en) | Object re-recognition method and apparatus, electronic device, program, and storage medium | |
| CN109446889B (en) | Object tracking method and device based on twin matching network | |
| US20240161461A1 (en) | Object detection method, object detection apparatus, and object detection system | |
| CN109543606A (en) | A kind of face identification method that attention mechanism is added | |
| CN108229355A (en) | Activity recognition method and apparatus, electronic equipment, computer storage media, program | |
| CN114118259B (en) | Target detection method and device | |
| CN111428743B (en) | Commodity identification method, commodity processing device and electronic equipment | |
| CN108229533A (en) | Image processing method, model pruning method, device and equipment | |
| CN113627411A (en) | Super-resolution-based commodity identification and price matching method and system | |
| CN113537397B (en) | Target detection and image definition joint learning method based on multi-scale feature fusion | |
| CN114359819A (en) | Image processing method, apparatus, device, storage medium and computer program product | |
| CN103793717A (en) | Methods for determining image-subject significance and training image-subject significance determining classifier and systems for same | |
| CN114548208A (en) | An improved real-time classification and detection method of plant seeds based on YOLOv5 | |
| CN114067275A (en) | Target object reminding method and system in monitoring scene and electronic equipment | |
| CN109977875A (en) | Gesture identification method and equipment based on deep learning | |
| CN111144220B (en) | Personnel detection method, device, equipment and medium suitable for big data | |
| Liu et al. | Generating underwater images by GANs and similarity measurement | |
| CN116958615A (en) | Image recognition methods, devices, equipment and media | |
| Bouafia et al. | SES-ReNet: Lightweight deep learning model for human detection in hazy weather conditions |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |